Cntx UI

Provides an MCP server to expose bundles and files for AI tools via MCP.

typescript

0

GitHub Stars

typescript

Language

6 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "nothingdao-cntx-ui": {
      "command": "cntx-ui",
      "args": [
        "mcp"
      ]
    }
  }
}

You can run cntx-ui as an MCP server to expose your project bundles and files to AI tools through the Model Context Protocol. This enables real-time access to bundles, file trees, and project status, so AI workflows can interact with your workspaces seamlessly.

How to use

Start and use the MCP server to expose project resources to AI clients. You can access bundles and files through MCP endpoints and use the built-in tools to inspect and regenerate bundles, view the file tree, and check project status. Real-time updates flow from the web interface to connected AI tools, making it easy to iterate across your bundles while you work.

Key MCP resources you will expose to AI clients include the following access points: cntx://bundle/<name> for XML content of a bundle and cntx://file/<path> for individual project files. Available MCP tools enable you to list bundles, retrieve a bundle, regenerate a bundle, obtain the project file tree, and query the current project status.

How to install

Follow these concrete steps to set up the MCP server and the surrounding tooling.

# Prerequisites
- Node.js >= 18.0.0
- npm

# Global installation (recommended)
npm install -g cntx-ui

# Local development setup
git clone https://github.com/nothingdao/cntx-ui.git
cd cntx-ui
npm install

# Optional: install web UI dependencies for the frontend
cd web
npm install
cd ..

Use the MCP server in practice

Initialize your project for cntx-ui if you haven’t already.
Start the MCP server in your project directory to enable AI tools to connect via MCP.
Open the web interface to configure bundles and monitor status, while keeping the MCP endpoint running for AI clients to query resources.

Project management with the web UI

Use the web interface to manage bundles, hidden files, cursor rules, and real-time updates. The UI provides controls to generate, inspect, and refresh bundles as you develop your AI workflows.

Available tools

list_bundles

List all available bundles exposed by the MCP server.

get_bundle

Retrieve the content of a specific bundle as XML.

generate_bundle

Regenerate a specific bundle to reflect current project state.

get_file_tree

Get the project file structure exposed through MCP.

get_project_status

Get the current status of the project, including bundle generation state and health.